Guessing

[JulianEdgar]

In another thread I wrote: "To be honest, I think the huge underlying issue - that I am afraid you, Freebeard are a strong part of - is that speculation has become largely the culture of this group. To be blunt, posters just guessing."(And I wasn't picking on Freebeard, but as a prolific poster he's a person who does a lot of guessing.)

What I meant by that is this.

Car aerodynamics is a very complex area. This is particularly so because changing one area of the car is likely to impact airflow behaviour in completely different areas. It is not like altering (say) suspension, where a change in suspension spring rates won't affect the engine's air/fuel ratio.

While I mention this idea in my Veloce book, after reading it, both Dr Thomas Wolf of Porsche and Rob Palin (ex Tesla) suggested to me that this point needs to be far more greatly emphasized. That is, the 'element reductionism' approach, where different aerodynamic elements of car can be viewed separately, is not a good idea. For example, Rob suggested to me that altering the rear separation height may impact the height of the front stagnation zone!

So any modification that makes a major change in airflow at one location (and if it isn't a major change, it is unlikely to do anything measurable) is likely to cause changes elsewhere.

Furthermore, the impact of a specific modification is very hard to assess based only on its description.

So - a simple example - a front air dam may increase or reduce drag. It depends on the existing undercar aero and the height of the air dam.

A rear spoiler may increase or reduce drag - on a car with attached flow, it depends on the increase in pressures of angled rear panels versus the increase in the size of the wake.

Boat-tail extensions may increase or reduce drag - it depends on the force vectors (direction and magnitude of the forces) as the airflow wraps around the curved panels versus the reduction in wake size versus the disruption in trailing vortices.

Obviously in that context, numerical / shape rules of thumb are, to put it mildly, highly problematic.

That's why the literature has no agreement on the best angle for rear diffusers, no agreement on the best angle for boat-tailing, no agreement on best ride height for low drag, no agreement on the ultimate low drag shape for a road car, no agreement on the best design for low drag wheels - and so on. (Those who pretend there is agreement simply haven't read widely and recently.)

Now does that mean that no advice can ever be offered?

No it does not.

For example, I think you can be pretty confident in saying that, on a car with a rough underside, a full undertray will reduce drag and lift. If you are building a car from scratch, any of the five(!) differently-shaped low drag templates in Hucho (2nd ed) would be good starting points. Wheels with flat faces are very likely to have lower drag than those with many exposed spokes.

But - and here's the key point - no-one can say with certainty what an outcome is until the modification has been tested. Without trialling modifications and testing them, you are - to a much greater degree than in other car modification - blundering around in the dark. But instead of saying that, on this group it's far more likely that someone will ask about a modification, and people will simply leap in with guesses - guesses that apparently are treated as quite credible.

But my observation is that those guesses are often based on rules of thumb, misquoted or selective use of references (eg quoted angles), irrelevant historic parallels (that aren't) and so on.

And let's look at that point about misusing examples from the past. When I have pointed out that an old reference may be useless as a guide to today's modification, it's been demanded of me: "Well, what in fluid dynamics has changed since that book was written?" The answer is: "Nothing."

But it's the wrong question.

Our engine designs don't look like those of the 1930s and 1940s because our understanding of combustion chamber design, tuned length intake systems, variable camshaft systems (etc, etc) has changed. As have our requirements of engines. We don't take the combustion chamber design of a Merlin and say that all modern engines should have their combustion chamber designs based on it - and if current engines aren't doing that, they're wrong. The combustion behaviour of fuels hasn't changed since the Merlin, but our understandings and requirements have.

On my YouTube channel people write in all the time asking for advice. So for example I run a video testing rear wing and spoiler combinations on the back of a Subaru Impreza for lift/downforce. Someone writes: "Is it the same with the DC5 Integra?". How would I know? How would anyone know who hasn't done testing (or seen manufacturer's testing)? How could such a question even be asked - or expect to be answered?

So the next time someone here asks for advice on a car aero modification, stand by - unfortunately - for the contribution of a lot of guesses.

[freebeard]

I'll respond to the bolded parts since they are presumably salient.

Car aerodynamics is a very complex area. This is particularly so because changing one area of the car is likely to impact airflow behaviour in completely different areas. It is not like altering (say) suspension, where a change in suspension spring rates won't affect the engine's air/fuel ratio.

Like fixing the Beetle too-fast back can impede engine cooling. Like that?

Furthermore, the impact of a specific modification is very hard to assess based only on its description.
[snip]
But - and here's the key point - no-one can say with certainty what an outcome is until the modification has been tested.

There exists received folk wisdom. Using, again, the Beetle as an example. 145s on the front, a rake, de-burblized (No side trim, fender-top turn signals, running boards). Collectively, Cal-Look. It was hashed out 1/4 mile at a time. Pro-street gave us the doghouse boat tail.

[JulianEdgar]

Originally Posted by freebeard

There exists received folk wisdom. Using, again, the Beetle as an example. 145s on the front, a rake, de-burblized (No side trim, fender-top turn signals, running boards). Collectively, Cal-Look. It was hashed out 1/4 mile at a time. Pro-street gave us the doghouse boat tail.

Isn't that testing? The very point I am making?

[freebeard]

Originally Posted by myself in another thread

So, we can agree to agree then?

If trolling aerohead is getting tired, I'm not available as the next in line.

[JulianEdgar]

Originally Posted by freebeard

So, we can agree to agree then?

If someone has done testing on a particular car and can demonstrate the positive results, fantastic!

But if we were to take the advice of what worked on the Beetle and apply it to a Prius - and you see stuff like that here all the time, just not with those models - then it is not much better than guessing.

That is my point!

[freebeard]

Fair enough. I gave it my best shot here:



Stock Prius with Moons and spats, and FWD aircooled VW internal air flow. Well... a Porsche fan, Golf front axles....

I'd expect a 36hp DSS class engine to have economy equal to the stock Prius with maybe less acceleration (less weight).

36hp Challenge Classes and Rules

DSS (Dual Super Stock)
Pre 1965 Dual or non Stock single barrel carburetor system. (Weber 48 IDA carburetors are pre 1965!) fitted to modified stock VW 36hp cylinder heads or period aftermarket heads (replicas like Wolfsburg West Okrasa heads are O.K.!). Handcrafted intake manifolds are O.K.! Engine must retain stock cylinder head stud locations (8mm studs are O.K.!). Requires any Bosch
distributor and coil. No camshaft or header limitations. Deck lid standoffs are NOT legal!

As long as I'm in the albums, here's me mocking the Template™



Actually, a centered drivers seat and you could lose the top and rear fenders. Hmmm...

[JulianEdgar]

Originally Posted by freebeard

Fair enough. I gave it my best shot here:

Very good!

[Grant-53]

What gets measured gets improved. I was well paid as a test /QA technician. That is why I do coast down tests and record the data. My hill drops 30 feet in 0.1 mile so I can compare different bikes.

[JulianEdgar]

Originally Posted by Grant-53

What gets measured gets improved. I was well paid as a test /QA technician. That is why I do coast down tests and record the data. My hill drops 30 feet in 0.1 mile so I can compare different bikes.

Yes, it's a major reason why I am looking forward to your results.

[jakobnev]

If you don't guess at all, you won't know which test to perform.